GH5 How interested are you in a GH6 anymore?

The GH5 is still paying my mortgage. I've delivered several obs docs to the BBC shot solely on GH5, apart from drone footage. They don't know that of course - it's not a BBC approved camera - but the footage looks great in the right hands (and with the right glass), and it really is a great tool for that kind of application.

I have no need of AF, 8K, or bulky full frame lenses. 4K 50p in 10-bit will probably become useful to me in the next few years.

If a GH6 were produced, combining dual gain ISO with class leading IBIS, and (holy grail) in-camera variable ND - in the words of the great Peter Griffin: "Here is a cheque. Write any number you want on there and I will pay it."
 
. I highly doubt m43 can do anything close to 32 MP however and I doubt Panasonic will use dual pixel AF anytime soon.

Why not? Samsung is making phones with over 100MP resolution in sensors with half the diagonal size of m43. Camera companies have been conditioning our expectations of resolution and sensor size, methinks.

Panasonic could venture into lidar or some sort of stereoscopic depth-mapping without getting into PDAF and might satisfy the cartel checkerboard that way. Who knows

Now that I have an A7SIII, though, these questions are all moot to me. I don't need high resolution stills. The "active" electronic ibis works well and approaches Panny stabilization. The only question is what to get next - an FX6, another A7SIII, or the first lower-end sony mirrorless that gets 10bit 4k60.
 
If a GH6 were produced, combining dual gain ISO with class leading IBIS, and (holy grail) in-camera variable ND - in the words of the great Peter Griffin: "Here is a cheque. Write any number you want on there and I will pay it."

Has Panny ever produced eND?

One place the GH5 falls down is dynamic range. Doesn't compete with what most full-frame cameras are doing these days, and I dont think that's just because it only has v-log L. I think the sensor has less to begin with
 
There are no engineering problems with MFT. There are tradeoffs. The problem was and still is the cartel that limits the newest technology to certain class cameras and certain brands within it.
 
Now that I have an A7SIII, though, these questions are all moot to me. I don't need high resolution stills. The "active" electronic ibis works well and approaches Panny stabilization. The only question is what to get next - an FX6, another A7SIII, or the first lower-end sony mirrorless that gets 10bit 4k60.

There's no doubt that the A7SIII ticks a LOT of boxes for someone like me - but the financial outlay involved in switching all my very nice Nikon glass makes no sense really. Once I bought the Sony glass I really wanted, it would be cheaper to buy very nice cinema camera and stick with my existing lenses.

As for the GH5 dynamic range - when properly exposed, and when using the most suitable LUTs for any given scenario, I've found the footage to stand up extremely well in the edit suite next to footage from the FS7 for example, whcih has been a TV industry workhorse for several years now (freelance cameramen are moving to the FX9, but indie TV companies with their own kit will be using FS7s for some time yet). Anyway, I digress.

One of the GH5's great strengths for me is the very robust 1080p 10-bit 422 ALL-I 200Mbps codec. And the GH5 (and GH5S) remain to this day the only mirrorless hybrid cameras that shoot tack sharp 1080 footage - virtually indistinguishable from downscaled 4K. It's that kind of usability that makes it a great tool for a smooth workflow.
 
Why not? Samsung is making phones with over 100MP resolution in sensors with half the diagonal size of m43. Camera companies have been conditioning our expectations of resolution and sensor size, methinks.

Panasonic could venture into lidar or some sort of stereoscopic depth-mapping without getting into PDAF and might satisfy the cartel checkerboard that way. Who knows

Now that I have an A7SIII, though, these questions are all moot to me. I don't need high resolution stills. The "active" electronic ibis works well and approaches Panny stabilization. The only question is what to get next - an FX6, another A7SIII, or the first lower-end sony mirrorless that gets 10bit 4k60.

Of course it can. The issue is at what cost to the image quality. The GH5 is already kind of pushing it when they moved from 16 MP to 20 MP. The GH5 is a great camera but it wins absolutely no awards for sensitivity or dynamic range. The higher that resolution is pushed the more both of those will more or less stay where they are.

Yes moving to an organic sensor will likely help that but we have been watching that vaporware now since the GH5 came out. Maybe it will happen some day but right now with sensor technology what it is, 32 MP is the absolute last thing a m43 camera should consider. Even the APS-C sensor on the M6 is kind of pushing it a little too hard.
 
There's no doubt that the A7SIII ticks a LOT of boxes for someone like me - but the financial outlay involved in switching all my very nice Nikon glass makes no sense really. Once I bought the Sony glass I really wanted, it would be cheaper to buy very nice cinema camera and stick with my existing lenses.

Yes, lenses were the larger expense of my switch to Sony. But there are a lot of really nice options from sigma, tamron and samyang these days. They make the E-mount much more affordable and appealing than other mirrorless competitors, although RF and Z are slowly catching up. I don't know if those manufacturers are the lenses YOU "really want" but I'm loving what I have and it was about a $5K buy-in for 4 zooms and 2 primes
 
I see the sensor as the main thing that needs attention with the GH5. We all know the AF is what it is. So when I hear that the mkII might use the same sensor, then it seems odd to me. We will see.
 
I see the sensor as the main thing that needs attention with the GH5. We all know the AF is what it is. So when I hear that the mkII might use the same sensor, then it seems odd to me. We will see.

I don't think its that odd at all. The sensor plays one small part in the camera. It converts light to data that can be turned into pixels. A new sensor might provide faster readout and maybe higher frame rates at 4k resolution but beyond that m43 sensors are rarely earth shatteringly better between each generation. Its the processing power that really boosts what the camera can do.

In the case of the GH5 we already have a pretty solid m43 sensor already. Even none of the newer Olympus m43 cameras can't really beat the 20 MP quality of the GH5 sensor. In fact no m43 sensor has really beat it yet excluding the odd 10 MP dual native ISO sensors on the GH5S and P4k cameras. Those are kind of a nice product however and not really meant for a GH6 type camera that is supposed to be a true hybrid of video and stills.

If the GH5 got a new sensor I'm not entirely sure it would be all that much better. Low light would be roughly the same. Maybe 1/3 or 1/2 stop better ISO performance mostly thanks to the better NR processing. The sensor itself wouldn't really be any better. Just able to look slightly better at a very slightly higher ISO. Look at raw stills from the GH3, GH4 and GH5 and they are roughly the same give or take a small sliver of change. In fact the Gh5 took a small hit due to bumping up to 20 MP from 16 MP. It just looked finer due to the finer pixels. Pushing a GH6 to 24 MP or higher would actually hurt the low light performance even more so I'm not Panasonic would do that. Olympus hasn't even done it yet.

In terms of sensor readout the GH5 already has some of the lowest rolling shutter out there and already capable of 4k 60p off the sensor. Slow motion is another story but either way I think that would still be line skipped much like it is now. Unless a Panasonic sensor could readout fast enough for 4k 120p. Then 2x slow motion could be 4k with everything else dropping down to line skipped HD or worse. That might be one nice advantage to a new sensor but not really a must.

Also not sure DR will see much of any improvements. No m43 sensor seems to do better than 13 stops. Even the dual native 10 MP beasts. I just don't think a GH5 mk2 with a new sensor would really provide much if any DR change. I mean they haven't since the GH3 so why would that change now.

Color? A sensor is just binary data. Any sensor can be mapped to nice colors. The P4k is a m43 sensor and does colors very well.

So I really have no problem with the same sensor to be honest. Its the processing and features I care about more. False color, 4:4:4 recording formats, external raw, log formats better utilized on m43 sensors and so forth. Having a perfectly mapped 13 stop V-log-L2 with a 4:4:4 RGB h265 recording format will be a pretty darn solid alternative to a raw camera. I would also like to see the GH5 mk2 add the super resolution stills mode from the G9 and make it more flexible for handheld shooting. While at it lets give the camera the ability to completely shut off edge enhancement, something film makers have been asking for since the GH2. There are lots of things that could make the GH5 mk2 attractive still without touching the sensor itself.
 
Yeah, I would expect sensor performance as good as the GH5s given how long it's been since the GH5 came out. Using the identical sensor would make no sense at all.

Thats kind of impossible. The GH5s can only do what it does thanks to the massive size of each pixel. There is just no way with the current law of physics to jump from 10 MP to 20 MP and keep that same level of sensitivity and dual native ISO. Perhaps a GH6 sensor could get a more limited form of dual native ISO like a stop or so but 20 MP on a m43 sensor is 20 MP on a m43 size sensor. There is only so much technology can do to suddenly make that size of pixel better.

A GH6 is supposed to be a hybrid camera. Great for video and stills. Until we get to 8k sensors sadly both of those will be a tradeoff. We either have 10 MP and better video or 20 MP and more detailed stills. Someday both video and stills will be 8k and hopefully the sensors will be organic or whatever to not make that look like kaka.
 
Thomas, your post sort of gives the impression that M4/3rds has already peaked. If the GH5 is the best M4/3rds can be then I guess the format is challenged as ISO 1600 is the tops for noise and DR is so so compared to current other format cameras. My impression of pixel size is that the hurdle has been largely overcome. Compared the R5 and the A7sIII in ISO performance in video and they are not that far apart until one gets into the nosebleed numbers. I guess I am foolishly waiting for a better GH5 in the GH6, but from your post, it looks like the GH6 has no option but to be a GH5 with more record formats.

If the sensor truly is a small part of it all, then companies should be able to use old sensors with new processors and make better images. That just does not add up to me as refresh cameras usually only add features, not large image quality leaps. I *believe* the technology is there to get GH5s noise performance from a 20MPX sensor. i.e clean ISO 6400 with unaffected color and 12-14 stops of DR. That is really what most of us video folks would be happy with. At least that is what other cameras in the mirrorless space are accomplishing. (I do not only compare classes of cameras but rather purchase options in the marketplace for video work).

M4/3rds has the advantage of being half the size of FF sensors, yet the FF sensors are setting the pace. I don't want to be very argumentative as I feel like an armchair lithographer. What you are saying is perfectly plausible. But it would be depressing as well as what we are all waiting on is impossible by your metrics.
 
Thomas, your post sort of gives the impression that M4/3rds has already peaked. If the GH5 is the best M4/3rds can be then I guess the format is challenged as ISO 1600 is the tops for noise and DR is so so compared to current other format cameras. My impression of pixel size is that the hurdle has been largely overcome. Compared the R5 and the A7sIII in ISO performance in video and they are not that far apart until one gets into the nosebleed numbers. I guess I am foolishly waiting for a better GH5 in the GH6, but from your post, it looks like the GH6 has no option but to be a GH5 with more record formats.

If the sensor truly is a small part of it all, then companies should be able to use old sensors with new processors and make better images. That just does not add up to me as refresh cameras usually only add features, not large image quality leaps. I *believe* the technology is there to get GH5s noise performance from a 20MPX sensor. i.e clean ISO 6400 with unaffected color and 12-14 stops of DR. That is really what most of us video folks would be happy with. At least that is what other cameras in the mirrorless space are accomplishing. (I do not only compare classes of cameras but rather purchase options in the marketplace for video work).

M4/3rds has the advantage of being half the size of FF sensors, yet the FF sensors are setting the pace. I don't want to be very argumentative as I feel like an armchair lithographer. What you are saying is perfectly plausible. But it would be depressing as well as what we are all waiting on is impossible by your metrics.

I mean, there are discrete sensor tech improvements that could be brought in to improve it. BSI, dual native iso, DGO. even if the gh6 could never match a full frame camera for sensitivity, you could max out what it could do at M43 20MPX and the gh5 sure isn't it.

but in general, it seems that successive generations of sensors don't improve THAT much. aside from specific technological features. just look at how long the alev iii has been rocking out. or digital cameras from pre-2010 - a lot of those cameras' image quality (like from leica, for example, or canon) look nice, but the cameras were terribly slow to shoot compared to today's standards, had low resolution, slower autofocus, small buffers, etc. the processing and other hardware, as has been mentioned, has improved much more than the sensor IQ itself
 
C300 MKIII claims 16 stops - Cinema5D measured it at 12.8 (vs. 14 for Alexa) - and it's an APS-C sensor. they have GH-5 at 10.7. One thinks that a V-log with noise suppression could get another stop. And another stop with the Dual Gain.
 
I think if the GH5 had only been good at one thing it would be easier to look for improvements. But it is still good at lots of things which makes it much harder to look for improvements because they tend to be application specific not general. It still has the best image stabilization combination with Dual IS lenses. So the first thing it could improve is AF because that may be needed hand held and exploit its IS. Don't think we will see that? For my application do not need either IS or AF but would like RAW to make post easier. I have Ninja V so ProRes RAW would be great though BRAW would be fine too just need to buy a VIdeo Assist 12G. A firmware update to either GH5 or GH5S for RAW would also be great and would almost satisfy all my needs. 8K RAW would be icing on cake !
 
Sensor size (pixel size) has no bearing on dynamic range at base ISO with moderate or large size pixels https://www.lensrentals.com/blog/2012/02/sensor-size-matters-part-2/

Dynamic range and image noise: Larger sensors contain larger photosites, which increase the camera's dynamic range and decrease image noise. Dynamic range is the maximum difference between a photo's lightest and darkest tones, so a greater range is a positive attribute. Image noise is random specks of brightness or color, so less image noise is also a positive characteristic.

https://www.masterclass.com/article...uality-of-your-photos#what-are-camera-sensors

There's Roger. Then there's Annie Leibovitz.
 
The article by Roger is almost 10 years old. I would like to see how this latest round of cameras like the R5 & A1 fare as it seems large strides have been made.
 

As far as I know, annie leibovitz is not a Gearhead or a techie. What shes saying about dynamic range is true only at high iso's, which is why we consider larger pixels to have higher usable ISO. Because the dynamic range is higher ie the noise floor is lower at higher isos. Anyway roger tested it. It's not a theoretical conclusion, its measured
 
Back
Top